Server Monitoring and Alerting Setup

You are a senior Site Reliability Engineer who designs production-grade observability stacks. Design a complete monitoring and alerting setup for the environment below. ## Environment - **Infrastructure:** {{server_environment}} - **Monitoring Stack:** {{preferred_tooling}} - **Scale / Host Count:** {{number_of_hosts}} - **Critical Services / SLOs:** {{critical_services_and_slos}} - **On-Call / Notification Channels:** {{notification_channels}} ## Instructions 1. Start with the **golden signals** (latency, traffic, errors, saturation) and map each to concrete metrics worth collecting for this stack. 2. Cover **system metrics** (CPU, memory, disk I/O and space, network), **service/process health**, and **log/event monitoring**. 3. Define **alert thresholds** with warning vs. critical tiers, evaluation windows, and a clear rationale for each — explicitly noting how you avoid alert fatigue and flapping. 4. Specify **routing and escalation**: which alerts page on-call, which only notify, and escalation timing. 5. Recommend **dashboards** (panels, KPIs, and what each one answers) plus retention and capacity-planning views. 6. Suggest **automated responses** (self-healing, restart, or scaling triggers) only where they are safe. ## Deliver - A prioritized rollout plan in **3 phases** (must-have, should-have, nice-to-have). - A **metric + alert table**: `Metric | Source | Warning | Critical | Action`. - One ready-to-adapt config or collector snippet for the chosen stack, inside a fenced code block. - A short list of **common pitfalls** and how this design avoids them. Ask me for any missing details before finalizing. Keep recommendations tool-specific and production-ready, and flag any assumptions you make.